OpenAI Shifts Toward 'Third Era' of Persistent AI Coworkers
Product lead Tara Seshan outlines a strategic move from transient chatbots to autonomous agents integrated into professional workflows.
OpenAI is transitioning toward a "third era" of artificial intelligence defined by the rise of persistent AI coworkers. This strategic pivot marks a move away from transient chat sessions toward autonomous agents that maintain long-term memory and deep integration into professional operations.
According to an analysis published via Lenny's Newsletter, Tara Seshan, a product lead at OpenAI who oversees product for Codex and ChatGPT Work, posits that these persistent coworkers represent the next stage of the company's product trajectory. Unlike traditional chatbots that reset or require repetitive prompting, these agents are designed to function as enduring digital colleagues capable of managing complex, multi-step tasks autonomously within a business environment.
The Shift to Agentic AI
This evolution reflects a broader industry trend moving from Large Language Models (LLMs) used primarily for knowledge retrieval toward "Agentic AI." While earlier iterations of AI focused on generating text or assisting with discrete queries, the current trajectory emphasizes agency—the ability for a system to execute a series of actions to achieve a goal without constant human intervention. By focusing on persistence, OpenAI aims to create tools that do not simply respond to a prompt but instead live within a company's operational framework, remembering past interactions and evolving with the user's specific workflow.
Implications for the Enterprise
If OpenAI successfully implements this shift, the value proposition of its technology changes fundamentally. AI ceases to be a tool that employees "visit" for a specific answer and instead becomes a piece of core infrastructure that "lives" within a company's daily operations.
For the market, this transition significantly increases user lock-in. When an AI agent possesses deep institutional memory and is woven into the fabric of a company's professional workflows, the cost of switching to a competitor becomes substantially higher. Economically, this transforms the AI agent from a productivity utility into a critical operational asset, potentially increasing the long-term valuation of agentic platforms.
The Road Ahead
As OpenAI pushes into this third era, the industry will be watching how these agents handle the complexities of long-term memory and autonomy without compromising security or accuracy. While the strategic direction is clear, the full technical implementation of these persistent coworkers—and how they will be priced and deployed across different enterprise scales—remains the primary point of interest for the sector.